Skip to content
ML Academy

Business judgment

Interview question

  • midlevel
  • 10 minsuggested time
  • 2Data scientist, ML engineer

A team proposes a forecasting project. Hypothetical estimates: build cost 250k, run cost 90k per year, benefit of 400k per year if it works, with a ramp to half the benefit in year 1, and a 60% chance the model reaches the required accuracy. Over three years, is it worth doing, and what would change your mind?

Tests: Expected-value reasoning for an ML project, risk-adjusting benefits, discounting, break-even probability, and communicating assumptions rather than a single number.

day 1day 16

Sign in to practise: answers are saved before model answers are shown.